Leslie Ballas of Cedars-Sinai will present the phase II multi-institutional study of ultralow-dose (4 Gy) response-adapted radiotherapy for symptomatic bone metastases from multiple myeloma on September 27 at 2:00 pm ET in the Grand Ballroom, during the ASTRO 2026 Clinical Trials session.
The planned interim analysis was reported in 2023, so the direction of the result is not in doubt. What the full dataset must establish is something the interim analysis could not: whether a strategy of minimal initial dose with conditional escalation is preferable to a fixed conventional dose, and what the reduction actually buys.
What the trial tests
Patients with a known diagnosis of multiple myeloma and a painful bone lesion received a total dose of 4 Gy, delivered either as two fractions of 2 Gy or as a single 4 Gy fraction. Lesions at the base of skull, lesions requiring stabilization, lesions causing cord compression, and lesions previously irradiated or surgically treated were excluded. Reirradiation with a clinician-selected regimen could be considered at four weeks or later for indeterminate pain response or for pain progression.
The primary objective was to determine whether pain reduction at four weeks was comparable to historical controls. Pain was measured with the Brief Pain Inventory before treatment and at two, four and eight weeks and six months, and response was classified according to the international consensus criteria for palliative radiotherapy endpoints, incorporating both change in Brief Pain Inventory score and oral morphine equivalent dose. Secondary objectives included quality of life, analgesic reduction, time to and duration of relief, response in patients with more than one index lesion, and pain response after a second course where required. Changes to systemic therapy and bisphosphonates in the four weeks before and after radiotherapy were permitted but recorded.
The single-fraction option deserves emphasis, because it removes the argument the trial's own rationale leads with. Ultralow-dose palliation has been advocated principally on grounds of convenience and reduced interference with the scheduling of systemic therapy. Against 20 Gy in five fractions or 30 Gy in ten, that argument holds. Against 8 Gy in a single fraction, which is the evidence-based standard for painful bone metastases and is in wide use in myeloma, it does not hold at all. One visit is one visit. What is at stake is therefore not treatment burden but whether 4 Gy is sufficient, and whether halving the dose delivers a benefit that can be measured.
The interim data
The planned futility analysis after 40 patients reported a complete response in 48 percent, a partial response in 38 percent, and an indeterminate response in 13 percent, with one patient withdrawing. Overall pain response was 86 percent. Seven patients, 18 percent, requested reirradiation at four weeks or later. The futility threshold was not met and the Data and Safety Monitoring Committee recommended continuation.
Two secondary observations are more informative than the headline rate.
The magnitude of relief was substantial in responders. Median Brief Pain Inventory score fell from 3.75 to 0 in complete responders and from 4.00 to 1 in partial responders, with a median change of minus 3.25 across all responders. Patients with an indeterminate response began at 5.25 and ended at 4.75, which is to say they did not improve at all.
Median planning target volume differed monotonically across response categories: 81 cc in complete responders, 140 cc in partial responders, and 226 cc in those with an indeterminate response. Combined with the higher baseline pain scores in the indeterminate group, this points at disease burden rather than intrinsic radioresistance as the discriminating variable. With 40 patients the observation is hypothesis-generating and should not be used for selection, but it is the most actionable signal the interim analysis produced, and it aligns with retrospective data suggesting that patients with a higher lesion burden may require an equivalent dose in 2 Gy fractions of 20 Gy or more for local control.
The case for minimizing dose in myeloma
The conventional palliative dose in myeloma was never established by randomized comparison. Practice was inherited from solid-tumor bone metastasis trials, in which overall pain response is closer to 70 percent and no dose-response relationship has been demonstrated across the tested range. Myeloma is considerably more radiosensitive, with response rates above 80 percent, and its radiobiology resembles that of indolent lymphoma, where 4 Gy in two fractions has an established role and has been extended into response-adapted definitive protocols. Retrospective series have reported effective pain control at very low doses in uncomplicated osseous myeloma lesions, providing proof of concept for the prospective trial.
There is a further argument that has strengthened considerably since the trial was designed. Myeloma is now treated with therapies whose efficacy depends on the patient's own hematopoietic and immune reserve: autologous stem cell collection, chimeric antigen receptor T-cell products requiring lymphapheresis, and bispecific antibodies acting through endogenous T cells. Myeloma bone lesions are by definition in marrow-bearing bone, and radiation-induced lymphopenia is related to dose and irradiated volume.
That argument should not be overstated. A reported series of bridging radiotherapy before CAR-T therapy used a median dose of 22 Gy in three to eight fractions, with a median interval of 25 days to cell infusion, and found no increase in cytokine release syndrome or neurotoxicity relative to patients who received no bridging radiotherapy. Conventional palliative doses have not been shown to compromise immune-based therapy. The marrow argument is biologically coherent and clinically untested, and this trial will not test it unless it collected the relevant data.
What the full analysis should be examined for
Reirradiation as a primary measure of adequacy. In a response-adapted design the reirradiation rate is not a footnote; it defines the regimen. If the mature rate approaches the interim 18 percent, the strategy delivers 4 Gy to roughly four of five patients and a second clinician-chosen course to the remainder. If it rises materially with longer follow-up, the regimen becomes 4 Gy plus a conditional supplement, and the comparison against a single 8 Gy fraction weakens accordingly. The timing of reirradiation matters as much as the rate.
Durability at six months rather than response at four weeks. The primary endpoint is a four-week measurement in a population whose median survival is now measured in years. The six-month Brief Pain Inventory data, the duration of relief, and time to any subsequent intervention at the index site are what determine whether 4 Gy is adequate treatment or deferred treatment.
Whether the volume signal holds. If the planning target volume gradient persists in the full cohort, it offers a prospectively usable criterion for deciding who should not start at 4 Gy. That would make the strategy substantially more efficient and would be a more durable contribution than the overall response rate.
Attribution. Systemic therapy can itself relieve bone pain rapidly, and protocol-permitted changes around radiotherapy were recorded rather than prohibited. The consensus response criteria, which incorporate analgesic consumption, mitigate this but do not remove it. How concurrent systemic therapy changes are reported will affect how much of the response can be attributed to 4 Gy.
Whether anything marrow-related was captured. Lymphocyte counts, subsequent stem cell collection, or receipt of immune-based therapy would substantially strengthen the case for dose minimization. Their absence, which would be unsurprising in a protocol written before bispecific antibodies entered routine use, should be stated rather than inferred.
Boundaries of application
The exclusions define the regimen as precisely as the dose does. Cord compression is the clearest limit: short-course radiotherapy has been shown to be suboptimal for cord compression from myeloma, and nothing here addresses that setting. Lesions requiring stabilization are excluded because analgesia and structural integrity are different objectives. Skull base disease is excluded, and lesions previously irradiated or operated were not eligible, so the regimen is not validated in the retreatment setting despite its low dose making retreatment mechanically straightforward.
This is a trial of uncomplicated painful myeloma bone lesions, and any adoption would need to preserve that boundary explicitly.
Assessment
The most interesting feature of this study is structural rather than dosimetric. It tests a model of palliation in which the initial dose is set at the minimum that works for most patients and escalation is reserved for those in whom it demonstrably did not. That model is standard in indolent lymphoma and almost absent from palliative bone practice, where the dose is chosen once and revisited only on failure that the treating physician happens to learn about.
The limitation is that a single-arm phase II against historical controls can show that a lower dose is sufficient for analgesia. It cannot show equivalence to 8 Gy in a single fraction, and it cannot quantify what the reduction preserves. Since the single-fraction option removes the convenience differential entirely, the case for replacing 8 Gy rests on a benefit that has not yet been measured in this disease.
Expected findings
We anticipate a four-week pain response in the full cohort broadly consistent with the interim 86 percent, likely somewhat lower as the population broadens, and a reirradiation rate in the region of 18 to 25 percent. We expect the six-month durability data to be the most closely examined part of the presentation, and we expect the planning target volume relationship to be reported and appropriately hedged.
Those results would justify a randomized comparison of 4 Gy against 8 Gy in a single fraction, with pain response as one endpoint and marrow and lymphocyte outcomes as another. They would not justify replacing the current standard, and the distinction should survive the presentation intact. The most useful version of this result is not that 4 Gy suffices for everyone. It is that it suffices for most, and that the trial can identify in advance the patients for whom it will not.
We will report on the presentation from ASTRO 2026.
This is the third article in the RadOncToday series previewing the Clinical Trials and Plenary sessions at ASTRO 2026 over the three weeks before the meeting.